LONDON, April 12 (Reuters) - South Korea will host the

second global AI (artificial intelligence) safety summit on May

21-22, six months after Britain held the inaugural event at

Bletchley Park.

In a statement published on Friday, government officials

said the event would build on the "Bletchley Declaration", an

agreement between countries - including the U.S. and China - to

cooperate on AI safety.

CONTEXT

When Microsoft ( MSFT )-backed OpenAI released ChatGPT in

late 2022, many were impressed, while some were alarmed.

Former Google researcher and "godfather of AI" Geoffrey

Hinton said the technology posed a more urgent threat to the

world than climate change. Elon Musk called for an immediate

pause in the development of such systems.

In November last year, Britain hosted the first AI Safety

Summit, bringing together world leaders, tech moguls and

academics to discuss how best to regulate the rapidly developing

technology.

After the two-day event, more than 25 government

representatives signed a joint declaration agreeing that

countries would work together and establish a common approach on

oversight.

In order to keep pace with the technology, it was agreed

that future summits would take place every six months. South

Korea will hold a largely virtual event in May, with France

hosting the next in-person summit later in 2024.

KEY QUOTES

"The AI Safety Summit held in Bletchley last year marked a

milestone in which the government, industry and academia came

together to address the potential risks of AI, and I am pleased

that the Republic of Korea takes the baton from the UK," said

Lee Jong-Ho of South Korea's ministry of science and information

and communication technology.

"AI is advancing at an unprecedented pace that exceeds our

expectations, and it is crucial to establish global norms and

governance to harness such technological innovations to enhance

the welfare of humanity."